Xen project Mailing List

[Xen-users] VGA passthrough with Xen 4.3 and xl toolstack - performance degradation resolved?

To: "xen-users@xxxxxxxxxxxxx" <xen-users@xxxxxxxxxxxxx>

From: "H. Sieger" <powerhouse.linux@xxxxxxxxx>

Date: Tue, 11 Mar 2014 11:01:43 -0700 (PDT)

Delivery-date: Tue, 11 Mar 2014 18:05:30 +0000

Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=X-YMail-OSG:Received:X-Rocket-MIMEInfo:X-Mailer:Message-ID:Date:From:Reply-To:Subject:To:MIME-Version:Content-Type; b=odlS3j0wfs+qwhJhJ3l1yfo3L4b6tn+LZ745eecXI3Z3eASLuXx9g8cYpY8qc/1w/OCsJ2jXqde6yb9ur1TWpAU3v1SLVqVZ/MUF3y6GqJseUYZCNpnHRyWOdXbR7SzMiWb+llE1sZ8N3jSvDXBNHNu5KlbtcdHzuuzZrkWC1mk=;

List-id: Xen user discussion <xen-users.lists.xen.org>

Hello,

Hope you can help. A while ago users noted performance degradation or dom0 stability issues when shuting down a HVM guest that uses VGA passthrough (e.g. Windows 7), and booting up the guest again. A workaround was to eject the graphics card within Windows, before shutting down the guest. This process is described here: http://blog.ktz.me/?p=219.

I tried to follow those instructions, but my Windows 7 guest wouldn't boot again after installing the startup and shutdown scripts. Manual eject/insert of the graphics card running the "logon" script worked fine, however (i.e. not a problem with incorrect PCI device ID).

Following this failure and after removing the startup/shutdown scripts, I booted and shutdown the Windows guest repeatedly, each time running a Unigine benchmark to test graphics performance. Surprise: everything works just fine.

Now the question: I'm using a Nvidia Quadro 2000 card and I was wondering if this guest reboot issue was only related to AMD graphics cards?

Or has this bug been fixed in Xen 4.3? I like to know.

Thanks in advance.

_______________________________________________ Xen-users mailing list Xen-users@xxxxxxxxxxxxx http://lists.xen.org/xen-users